Short-term wind-power power forecasting method based on coding/decoding long-short term memory network

Abstract

The invention discloses a short-term wind-power power forecasting method based on a coding / decoding long-short term memory network. The method comprises steps of firstly, using an E-D based LSTM network to carry out AE processing on power, and extracting a trained network middle state as abstract expression of a time sequence relation in WP data; then, combining the network middle state extractedin the first step with weather data in a prediction period, and inputting the result into a new LSTM network, thereby finishing prediction of wind-power power. Compared with a multi-layer LSTM networkmethod where AE preprocessing is not used, according to the invention, by use of the WP time frequency relation information extracted in the AE process, the model misspecification risk of a model isreduced; the generalization ability is improved; and by combining the time sequence characteristic and the weather prediction information, the prediction precision is further improved.

Description

technical field [0001] The invention relates to the technical field of wind power prediction algorithms, in particular to a short-term wind power prediction method based on encoding and decoding long short-term memory networks. Background technique [0002] Wind energy is an important resource in clean and renewable energy, but due to the intermittent and random characteristics of wind energy, WP is uncertain and weakly controllable. This has brought hidden dangers and challenges to the safe and stable operation of the power grid. Accurate WPP can alleviate the pressure of power grid frequency regulation and peak regulation, and is of great significance for large-scale wind power grid integration and operation management. [0003] At present, the methods for wind power forecasting are divided according to the time period, mainly including: the long-term forecasting method based on the year; the medium-term forecasting method based on the month and week; the short-term forec...

AI Technical Summary

Problems solved by technology

Among them: the extrapolation method has relatively strict assumptions on the random distribution characteristics of the data; the kernel function selection of SVM is random, and the increase in the amount of sample data and the increase in the dimension of the input data will lead to higher computational complexity; shallow ANN Although it can fit the sample data relatively well, it has the disadvantages of overfitting and poor generalization ability.

Images